Scramble

Scramble is a python module designed to address the issue of heterogeneous N values in machine learning model classes. It provides a cross-validation method to ensure accurate model accuracy estimation, particularly for object detection and classification in satellite imagery. Scramble was developed to fill a gap in verification and validation methods for models with varying sample sizes in different classes. By randomizing and splitting data sets, Scramble allows for more reliable accuracy estimation during the training and testing process.

GitHub repository: https://github.com/IdahoLabResearch/Scramble